It’s started in TFA

You can't expect someone to read 4 paragraphs into an article before commenting

@grok is this true?

Sorry, @grok is offline after declaring himself MechaMussolini earlier today

It's a trade off.

For local LLM there are Ollama and LM Studio. How is this different?

This seems like a great question to ask an offline llm that runs on your mobile device. Does ollama and Lm studio run on your mobile device?

I was at an antique shop in Alameda about to buy a pulp sci-fi book. It was just a couple of bucks, nothing expensive. But the card reader wasn't accepting my Apple Pay. Another guy shopping there offered to pay for me. I told him I could Venmo or PayPal him the money, but he wouldn't take it. It wasn't a big deal in terms of money, but he didn't have to do that

My brother told me, many years ago "if you can do something nice for someone for a buck, you probably oughta just do it."

I didn't learn much from him, but that was a good one.

where do you live?

it depends of the sample. usually you have at least 1200, with a guaranteed of at least 800, so maybe he could ask for a refund.

yes, the difference here is that the $1000 tag is "at-scale price". You reach that price point by running multiple sequencing with a set of reactive.

Which data uses for training?

karpathy/fineweb-edu-100b-shuffle: https://huggingface.co/datasets/karpathy/fineweb-edu-100b-sh...

Which is derived from HuggingFaceFW/fineweb-edu: https://huggingface.co/datasets/HuggingFaceFW/fineweb-edu

HuggingFaceTB/smol-smoltalk: https://huggingface.co/datasets/HuggingFaceTB/smol-smoltalk

And extra fine-tuning on portions of:

cais/mmlu: https://huggingface.co/datasets/cais/mmlu

openai/gsm8k: https://huggingface.co/datasets/openai/gsm8k

allenai/ai2_arc: https://huggingface.co/datasets/allenai/ai2_arc

I think he mentioned somewhere he used fineweb (I assume this one https://huggingface.co/datasets/HuggingFaceFW/fineweb)

Shameless plug: made a parquet conversion utility: pip install parquetconv

It is a command line wrapper to generate a Pandas SF and save it as CSV (or the other way around)

"...the investigator thought that Tesla was being collaborative with the investigation at the time..."

So this is also a failure of the investigator.

No its not. Investigators will try to obtain data willingly unless they are made to believe that the party is not acting in good faith.